In many optical systems and applications, such as near-field microscopy, imaging, photolithography and optical storage it is important to reduce the spot size and thus obtain higher definition or resolution. The spot size of an optical system, e.g., an optical storage system, is commonly defined as the distance between half power points. This distance is determined by diffraction to be approximately .lambda./(2.multidot.NA), where .lambda. is the free space wavelength of the light used and NA is the numerical aperture of the objective lens focusing the light beam. NA is defined as NA=n sin(.theta.), where .theta. is the half cone angle of the focused light rays and n is the index of refraction of the medium in which .theta. is measured.
One way to improve the definition is to work at shorter wavelengths .lambda., e.g., in the green or blue range, and to increase the numerical aperture to be as close to one as possible. A further possibility is to employ near-field optics in the manner described by Betzig et al. in Applied Physics Letters, Vol. 62, pp. 142 (1992), using a tapered fiber with a metal film with a small pinhole at the end. The definition of the system is determined by the size of the pinhole, and can be 50 nm or less. The advantages of the fiber probe system are its excellent definition and its polarization preserving capability which is particularly useful in magneto-optic storage applications. The disadvantages of the system are its poor light efficiency and the fact that it can only observe a single spot at a time, thus limiting its tracking ability when used for optical storage.
Another alternative is to use a solid immersion lens (SIL) between the objective lens and the illuminated object, e.g., an optical recording medium or sample under investigation. The SIL is placed within a wavelength .lambda. or less (in the near-field) of the object. Optical systems taking advantage of appropriate SILs are described, e.g., by S. M. Mansfield et al. "Solid Immersion Microscope", Applied Physics Letters, Vol. 57, pp. 2615-6 (1990); S. M. Mansfield et al. "High Numerical Aperture Lens System for Optical Storage", Optics Letters, Vol. 18, pp. 305-7 (1993) and in U.S. Pat. No. 5,004,307 issued to G. S. Kino et al. In this patent Kino et al. teach the use of a high refractive index SIL having a spherical surface facing the objective lens and a flat front surface facing an object to be examined. The use of this SIL enables one to go beyond the Rayleigh diffraction limit in air. In one embodiment, the SIL is employed in a near-field application in a reflection optical microscope to increase the resolution of the microscope by the factor of 1/n, where n is the index of refraction of the SIL.
A paper by G. S. Kino presented at the SPIE Conference on Far- and Near-Field Optics, "Fields Associated with the Solid Immersion Lens", SPIE, Vol. 3467, pp. 128-37 (1998) describes in more detail the principles of operation of two particular SILs. The first is a hemispherical SIL and the second is a supersphere SIL or a stigmatic SIL. The hemispherical SIL improves the effective NA of the objective lens by the refractive index n of the SIL and decreases the spot size by 1/n. The supersphere SIL increases the effective NA of the objective lens by the square of the refractive index n.sup.2 and obtains a focus at a distance a/n from the center of the supersphere, where a is the sphere's radius. The spot size is reduced by a factor of n.sup.2. The performance characteristics and theoretical limitations of both types of SILs are also discussed.
SILs have found multiple applications. For example, Corle et al. in U.S. Pat. No. 5,125,750 teach the use of a SIL in an optical recording system to reduce the spot size in an optical recording medium. These SILs typically have a spherical surface facing the objective lens and a flat surface facing an optical recording medium. The flat surface is in close proximity to the medium.
In U.S. Pat. No. 5,497,359 Mamin et al. teach the use of a superhemisphere SIL in a radiation-transparent air bearing slider employed in an optical disk data storage system. Lee et al. in U.S. Pat. No. 5,729,393 also teach an optical storage system utilizing a flying head using a SIL with a raised central surface facing the medium. In U.S. Pat. No. 5,881,042 Knight teaches a flying head with a SIL partially mounted on a slider in an optical recording system. This slider incorporates the objective lens and it can be used in a magneto-optic storage system. Finally, in U.S. Pat. No. 5,883,872 Kino teaches the use a SIL with a mask having a slit for further reducing the spot size and thus increasing the optical recording density in an optical storage system, e.g., a magneto-optic storage system.
The prior art SILs as well as the optical systems using them have a number of shortcomings. Hemispherical SILs suffer from back reflection problems. These degrade system performance, especially when the light source is a laser, e.g., a laser diode, and the back reflection is coupled back into the laser. Also, the ray reflected from the spherical surface and the ray reflected from the flat surface or from an object just below the flat surface are coincident. This gives rise to undesirable interference effects.
Superhemispherical SILs have reduced back reflection. However, they demagnify the image of the object by a larger factor than hemispherical SILs. For example, the demagnification of superhemispherical SILs in the axial direction is 1/n.sup.3. Because of this, the length tolerance for the superhemispherical SIL is very tight. Both the hemispherical and superhemispherical SILs increase the effective NA (NA.sub.eff) of the objective lens (for hemispherical SIL NA.sub.eff =NA.sub.objective.multidot.n; and for superhemispherical SIL NA.sub.eff =NA.sub.objective.multidot.n.sup.2. The maximum NA.sub.eff that can be obtained by either type of SIL is NA.sub.eff =n.
Hemispherical, superhemispherical and related SILs experience alignment problems because optical systems employing them require the use of a separate objective lens. This separate lens has to be accurately aligned with the SIL. In many optical systems alignment between these two lenses can not be easily preserved due to external influences (vibrations, stresses, thermal effects etc.). In addition, in systems where the number of parts is to be small, e.g., for weight and size reasons the objective lens is cumbersome.